Adding an example with CLUT CSM1/2 variants

Author: fjtrujy

CMakeLists.txt

@@ -352,6 +352,7 @@ if(NOT SKIP_BUILD_EXAMPLES)
 		bigtex
 		coverflow
 		clut
+		clutcsm
 		cube
 		fb
 		fhdbg

examples/clutcsm/clutcsm.c

@@ -0,0 +1,163 @@
+//  ____     ___ |    / _____ _____
+// |  __    |    |___/    |     |
+// |___| ___|    |    \ __|__   |     gsKit Open Source Project.
+// ----------------------------------------------------------------------
+// Copyright 2004 - Chris "Neovanglist" Gilbert <Neovanglist@LainOS.org>
+// Licenced under Academic Free License version 2.0
+// Review gsKit README & LICENSE files for further details.
+//
+// textures.c - Example demonstrating how the 2 different CLUT storage modes work.
+// CSMT1 and CSM1 are the two different CLUT storage modes. CSM1 is the default
+// mode, and is the mode that is used which requires to have the content swizzled
+// before uploading it to the GS. However, CSM2 is a mode that allows you to
+// upload the CLUT data directly to the GS, without having to swizzle it first.
+// Additionally it allows you to upload a list of CLUTs, and then switch between
+// them using the GS_SETCLUT command.
+//
+//
+
+#include <stdio.h>
+#include <malloc.h>
+
+#include <gsKit.h>
+#include <dmaKit.h>
+#include <gsToolkit.h>
+#include <gsInline.h>
+#include <gsTexture.h>
+
+#define CLUT_SIZE 256
+#define TEXTURE_WIDTH 320
+#define TEXTURE_HEIGHT 240
+
+#define USING_CSM1 0
+
+#if USING_CSM1
+#define TEXTURE_CLUT_WIDTH 16
+#define TEXTURE_CLUT_HEIGHT 16
+#else
+#define TEXTURE_CLUT_WIDTH 256
+#define TEXTURE_CLUT_HEIGHT 1
+#endif
+
+int main(int argc, char *argv[])
+{
+	GSGLOBAL *gsGlobal;
+	GSTEXTURE tex;
+	uint32_t i, j, offset;
+	uint32_t totalVertices = 2;
+	uint64_t White = GS_SETREG_RGBAQ(0xFF, 0xFF, 0xFF, 0x00, 0x00);
+	gs_rgbaq color = color_to_RGBAQ(0x80, 0x80, 0x80, 0x80, 0);
+
+	gsGlobal = gsKit_init_global();
+
+	gsGlobal->PSM = GS_PSM_CT24;
+	gsGlobal->PSMZ = GS_PSMZ_16S;
+
+	dmaKit_init(D_CTRL_RELE_OFF, D_CTRL_MFD_OFF, D_CTRL_STS_UNSPEC,
+				D_CTRL_STD_OFF, D_CTRL_RCYC_8, 1 << DMA_CHANNEL_GIF);
+
+	// Initialize the DMAC
+	dmaKit_chan_init(DMA_CHANNEL_GIF);
+
+	gsKit_init_screen(gsGlobal);
+	gsKit_mode_switch(gsGlobal, GS_ONESHOT);
+
+	tex.Width = TEXTURE_WIDTH;
+	tex.Height = TEXTURE_HEIGHT;
+	tex.PSM = GS_PSM_T8;
+	tex.ClutPSM = GS_PSM_CT16;
+	tex.Clut = memalign(128, gsKit_texture_size_ee(CLUT_SIZE, 1, tex.ClutPSM));
+	tex.Mem = memalign(128, gsKit_texture_size_ee(tex.Width, tex.Height, tex.PSM));
+	tex.Vram = gsKit_vram_alloc(gsGlobal, gsKit_texture_size(tex.Width, tex.Height, GS_PSM_T8), GSKIT_ALLOC_USERBUFFER);
+	tex.VramClut = gsKit_vram_alloc(gsGlobal, gsKit_texture_size(TEXTURE_CLUT_WIDTH, TEXTURE_CLUT_HEIGHT, GS_PSM_CT16), GSKIT_ALLOC_USERBUFFER);
+#if USING_CSM1
+	tex.ClutStorageMode = GS_CLUT_STORAGE_CSM1;
+#else
+	tex.ClutStorageMode = GS_CLUT_STORAGE_CSM2;
+#endif
+
+	printf("Tex VRAM Range = 0x%X - 0x%X\n", tex.Vram, tex.Vram + gsKit_texture_size(tex.Width, tex.Height, tex.PSM) - 1);
+
+	gsKit_set_clamp(gsGlobal, GS_CMODE_CLAMP);
+
+	uint16_t colors[4] = {
+		0xEFFF,
+		0x001F,
+		0x03E0,
+		0x7C00,
+	};
+
+	// fill the clut, each 16 positions are the same color
+	uint16_t *clut = (uint16_t *)tex.Clut;
+	for (i = 0; i < CLUT_SIZE; i += 16)
+	{
+		for (j = 0; j < 16; ++j)
+		{
+			clut[i + j] = colors[(i / 16) & 0x3];
+		}
+	}
+#if USING_CSM1
+	// Swap the texture to have the CSM1 requirements
+	for (i = 0; i < CLUT_SIZE; i++)
+	{
+		if ((i & 0x18) == 8)
+		{
+			uint16_t tmp = clut[i];
+			clut[i] = clut[i + 8];
+			clut[i + 8] = tmp;
+		}
+	}
+#endif
+
+	// print the clut, 16 colors per line
+	for (i = 0; i < CLUT_SIZE; i += 16)
+	{
+		for (j = 0; j < 16; ++j)
+		{
+			printf("%04X ", clut[i + j]);
+		}
+		printf("\n");
+	}
+
+	// initialize texture, each 16 horizontal pixels are the same color, so let's point to the right color
+	uint8_t *tex256 = (uint8_t *)tex.Mem;
+	for (i = 0; i < TEXTURE_HEIGHT; ++i)
+	{
+		for (j = 0; j < TEXTURE_WIDTH; ++j)
+		{
+			tex256[i * TEXTURE_WIDTH + j] = j & 0xFF;
+		}
+	}
+
+	gsKit_texture_send((u32 *)tex.Mem, tex.Width, tex.Height, tex.Vram, tex.PSM, 1, GS_CLUT_TEXTURE);
+	gsKit_texture_send((u32 *)tex.Clut, TEXTURE_CLUT_WIDTH, TEXTURE_CLUT_HEIGHT, tex.VramClut, tex.ClutPSM, 1, GS_CLUT_PALLETE);
+
+	GSPRIMUVPOINTFLAT *verts = (GSPRIMUVPOINTFLAT *)malloc(sizeof(GSPRIMUVPOINTFLAT) * totalVertices);
+	verts[0].xyz2 = vertex_to_XYZ2(gsGlobal, 0, 0, 0);
+	verts[0].uv = vertex_to_UV(&tex, 0, 0);
+
+	verts[1].xyz2 = vertex_to_XYZ2(gsGlobal, gsGlobal->Width, gsGlobal->Height, 0);
+	verts[1].uv = vertex_to_UV(&tex, tex.Width, tex.Height);
+
+	offset = 0;
+
+	gs_texclut texclut = postion_to_TEXCLUT(4, 0, 0);
+
+	while (1)
+	{
+		gsKit_clear(gsGlobal, White);
+#if !USING_CSM1
+		gsKit_set_texclut(gsGlobal, texclut);
+#endif
+		gskit_prim_list_sprite_texture_uv_flat_color(gsGlobal, &tex, color, totalVertices, verts);
+
+		gsKit_queue_exec(gsGlobal);
+		gsKit_sync_flip(gsGlobal);
+
+		offset++;
+	}
+
+	free(verts);
+
+	return 0;
+}